Meal timing, fasting and glucocorticoids interplay in serum leptin ...

184 A Elimam and C Marcus EUROPEAN JOURNAL OF ENDOCRINOLOGY (2002) 147
Diagnostics, Mannheim, Germany). Serum insulin and
cortisol concentrations were determined by RIA with
Pharmacia Insulin RIA (Pharmacia & Upjohn, Uppsala,
Sweden) and DSL Cortisol RIA (Diagnostic Systems Laboratories,
Webster, TX, USA) respectively.
Statistics
Data are presented as means^S.E., and statistical significance
was accepted at P # 0:05: Statistical analysis
was done on absolute values, while in the figures the
data are presented as percent change from baseline.
Statistical differences over time were calculated by
ANOVA for repeated measurements with post-hoc testing
using Tukey’s test where appropriate. The variables
were log transformed for the analysis when they were
not normally distributed. Data were analysed with the
program Statistica (Statsoft, Inc., Tulsa, OK, USA).
Results
Serum leptin increased significantly within 5 h
following each meal (P , 0:01; Fig. 2). In the groups
eating 12-hourly, leptin declined in the evening but
rose again in response to the evening meal. The rise
in leptin after the meal taken in the evening (36 h)
was greater than after the same meal taken in the
morning (P , 0:001; Figs 2 and 3). In those fasting
for 48 h, leptin decreased by 60% and showed no diurnal
variation. Leptin rise was higher following 48 h of
fasting when compared with the response on day 1 in
the same individuals or the response to the same
Leptin response to a standard meal
(% change)
www.eje.org
60
50
40
30
20
10
0
Meals at 0, 24, 36 and 48 h
Meals at 0, 24, 36 and 48h (DEX)
Meals at 0 and 48h
Meals at 0 and 48h (DEX)
meal on day 3 in the groups eating 12-hourly
(P , 0:02; Fig. 3).
DEX increased leptin concentrations in those eating
intermittently (P , 0:02; Fig. 2). The difference was
statistically significant at 2000 h on day 2, i.e. after a
total of 0.4 mg DEX. Thereafter, the differences between
the two groups were constant, i.e. no cumulative effect.
There was no difference in leptin between those taking
DEX and controls in the groups fasting for 48 h.
Glucose and insulin increased significantly following
each meal (P , 0:001; Figs 4 and 5). In the groups
fasting for 48 h, the rise in glucose and insulin in
response to the meal taken on day 3 was significantly
higher when compared with the response on day 1
ðP , 0:001Þ:
The diurnal rhythm for cortisol was maintained in all
groups (Fig. 6). The groups taking DEX tended to have
lower nadirs, but the difference was not statistically
significant.
Discussion
Meal at 0 Meal at 24h Meal at 36h (night) Meal at 48h
Meal timing
We investigated the effect of different durations of
fasting with and without a low dose of glucocorticoids
on the leptin diurnal variation and response to a standard
meal. In a previous study, we reported for the first
time a 4-fold increase in serum leptin in response to
energy intake, TPN, given after surgery in otherwise
healthy patients (28). In addition, we also reported
that leptin concentrations did not decrease in response
to fasting in the control group that had only saline
infusion for 24 h after the surgery (28). Our hypothesis
Figure 3 Mean^S.E. serum leptin response
to a standard meal given at 0, 24, 36 and
48 h. The leptin response was calculated by
subtracting individual leptin levels just before
eating a standard meal from the levels 5 h
after the meal.